Doctors should stop prescribing chloroquine and its sister drug, hydroxychloroquine, to prevent or treat COVID-19, with rare exceptions. That's the main conclusion of a recent article by four physicians.

It isn't only because there's little evidence that the drugs are effective; the doctors' concern is also about the fact that there is now a national shortage of these drugs for people with conditions the drugs effectively treat, primarily the autoimmune disorders rheumatoid arthritis and lupus. These patients are now having trouble getting their prescriptions filled.

After Trump spoke, prescriptions for these drugs rose 46-fold in 24 hours. Yet there's little evidence that these drugs are effective in treating COVID-19.

Hydroxychloroquine is a mainstay of lupus treatment, helping prevent both flare-ups and organ damage.

In late March, the U.S. Food and Drug Administration issued an emergency use authorization for chloroquine and hydroxychloroquine for the treatment of COVID-19. That helped fuel interest in the drugs. But in late April, the FDA issued a warning that hydroxychloroquine can cause heart rhythm disturbances that can lead to sudden death.

Yet the greatest factor leading to the shortage may be the President himself. On April 4, he publicly stated: “What do you have to lose. Take it, I really think they should take it.” During the next 24 hours, prescriptions for these drugs rose 46-fold.

Yet there's little evidence that these drugs are effective in treating COVID-19.

Randomized controlled trials are considered the gold standard in judging a drug's effectiveness. To date there have only been three such studies published.

All three tested hydroxychloroquine alone versus standard care in China. One showed no significant difference in viral clearance at 28 days, the second, no difference at seven days. The third reported some improvements in fever and cough as well as in chest scans.

Not only is this little evidence of hydroxychloroquine's effectiveness, the three trials enrolled 225, 62 and 30 patients, too few to provide reliable evidence even if they had shown greater effectiveness.

There are also the worries about heart problems. One reason the drugs have a good safety profile in lupus patients is that they are mostly young or middle-aged, with hearts generally healthier than those of older people. The risks of heart complications from hydroxychloroquine for patients with COVID-19 are much higher because they are typically older or have existing heart problems.

Even for lupus patients, researchers have been questioning the benefit of continued hydroxychloroquine usage in older patients because of evidence of eye damage that occurs over time.

It's understandable that people might want to try these drugs. The possible payoff is substantial, just like it is with lottery tickets. But the doctors who have been prescribing them and the people who have been clamoring for these meds are causing harm to other patients. And there's little evidence that they're doing themselves any good.

The article appears in The American Journal of Medicine.